A method and device for identifying line topology in a low-voltage area

Abstract

The present invention relates to a low-voltage substation line topology identification method and device, as well as a low-voltage substation power grid acquisition system. In the line topology identification method, the topology is identified based on the attenuation value between two meter boxes. First, branch identification is performed. Then, based on whether the signal strength attenuation between the meter boxes meets the signal strength attenuation condition, a signal strength attenuation-based identification method or a machine learning-based identification method is selected to identify the meter boxes under the branch. Finally, based on the results of the branch identification and meter box identification, the order of the meter boxes under each branch is identified. The identification method and device provided by the present invention can be based on the original low-voltage substation power grid acquisition system. Through calculation of relevant methods, the low-voltage substation line topology identification function can be realized without modifying the hardware system. The identification accuracy is high and the implementation cost is low. Compared with the substation line topology identification solution in the prior art, it is more suitable for large-scale promotion and application.

Description

Technical Field

[0001] The present invention relates to the technical field of substation line topology identification, and in particular to a low-voltage substation line topology identification method and device. Background Art

[0002] In the prior art, the commonly used methods for identifying the line topology of low-voltage substations include the following: (1) Manually inferring the substation line topology based on the on-site environment, archival information, etc. This method is the most primitive method, with low efficiency, requiring a lot of manpower, and low recognition accuracy, and is not suitable for widespread promotion. (2) By placing a GPS chip in the control module to achieve the purpose of line topology identification, this method requires the modification of STA and other modules, and the promotion cost is high. (3) Modifying the STA module and adding a carrier signal generating device to it so that the STA module can emit a specific signal to determine the line topology relationship of the substation. This method also requires the modification of STA and the promotion cost is high.

[0003] In summary, the low-voltage substation line topology identification solutions in the existing technology all require changes to the STA, which has high promotion and implementation costs and requires significant changes to the entire system. Summary of the Invention

[0004] Based on the above situation of the prior art, the purpose of the present invention is to provide a low-voltage substation line topology identification method and device. This method, based on the original low-voltage substation power grid acquisition system, can identify the substation line topology through the application of the topology identification method without adding additional hardware. Compared with the substation line topology identification scheme in the prior art, it has lower cost and is more suitable for large-scale promotion.

[0005] To achieve the above object, according to one aspect of the present invention, a method for identifying line topology in a low-voltage area is provided, comprising the steps of:

[0006] Perform branch identification based on the signal strength attenuation between meter boxes;

[0007] Determine whether the signal strength attenuation between meter boxes meets the signal strength attenuation condition. If the signal strength attenuation between meter boxes meets the signal strength attenuation condition, use a recognition method based on signal strength attenuation to identify the meter box under the branch. If the signal strength attenuation between meter boxes does not meet the signal strength attenuation condition, use a recognition method based on machine learning to identify the meter box under the branch.

[0008] Based on the results of the branch identification and the meter box identification, the order of the meter boxes under each branch is identified.

[0009] Furthermore, the branch identification is performed according to the signal strength attenuation between the meter boxes, including:

[0010] When the attenuation value between two meter boxes is ≤ s, the two meter boxes are considered neighbors;

[0011] The set A(A1, A2, ..., A1) of neighbors of each meter box in the statistical area is n ) and the number of elements in each set A set a(a1,a2,……,a n );

[0012] Analyze the set a, accumulate the same number of a in r, r1 represents the number of elements in the set a that are 1, r2 represents the number of elements in the set a that are 2...r 10 Indicates the number of elements in set a that are 10, r 11 Indicates the number of elements in set a that are greater than 10;

[0013] Calculate the ratio of multiple meter boxes:

[0014] rate=(n–r1–r2) / n;

[0015] When rate>r thd When rate≤r thd When s=s+1, repeat the above steps;

[0016] Take the intersection of the elements in each set A and get m sets B (B1B2,…,B m ), set B is the result of branch identification;

[0017] Among them, n is the number of meter boxes in the area to be identified, s is the attenuation value threshold, r thd It is the empirical value of the ratio of multiple meter boxes.

[0018] Furthermore, the determining whether the signal strength attenuation between the meter boxes meets the signal strength attenuation condition includes:

[0019] Assume s = 1, and calculate r1 and r2;

[0020] If 2*r2≤n–r1, the signal strength attenuation between meter boxes meets the signal strength attenuation condition;

[0021] If 2*r2>n–r1, the signal strength attenuation between meter boxes does not meet the signal strength attenuation condition.

[0022] Furthermore, the identification method based on signal strength attenuation is used to identify the branch meter box, including:

[0023] Obtain the set A of neighbors of each table box under the branch to be identified and the set a of the number of elements in each set A;

[0024] When an element greater than 1 appears in a, proceed to the next step; otherwise, s = s + 1, and repeat the above steps;

[0025] s=s+s j , and repeat the above steps;

[0026] Intersect the elements in each set A to obtain multiple sets B, where set B represents the identification of the table box under this branch;

[0027] Among them, s j To decay experience value;

[0028] Furthermore, the identification method based on machine learning is used to identify the branch meter box, including:

[0029] Find the eigenvalue of the multiple linear regression formula between two meter boxes;

[0030] Select multiple sets of characteristic values ​​of the same meter box and different meter boxes as models;

[0031] Calculate the Euclidean distance between the eigenvalue and each set of eigenvalues ​​in the model;

[0032] Sort the Euclidean distances and take the three smallest ones. When there are many groups of characteristic values ​​of the same box represented by the three Euclidean distances, the result to be identified is the same box.

[0033] Furthermore, the multivariate linear regression formula includes the following formula:

[0034]

[0035] Where U is the meter box voltage, I is the meter box current, the superscript is time i, the subscript is meter box 1 or meter box 2, X and R represent impedance, and the subscripts R and X represent the real and imaginary currents, respectively.

[0036] Furthermore, the identification of the order of the meter boxes under each branch includes:

[0037] Get the set A of each table box under this branch;

[0038] If a meter box outside the branch appears for the first time in set A, take the meter box corresponding to set A as the first meter box in the branch, mark it as P, and proceed to the next step; otherwise, s = s + 1, and repeat the above steps;

[0039] By comparing the attenuation values ​​of the meter box in P and other meter boxes in the branch, find the meter box closest to P and mark it as P;

[0040] If P is the last marked table box in the branch, then end, otherwise repeat the previous step.

[0041] According to another aspect of the present invention, a low-voltage substation line topology identification device is provided, comprising a branch identification module, a meter box identification method selection module, a meter box identification module, and a meter box sequence identification module; wherein,

[0042] The branch identification module performs branch identification based on the signal strength attenuation between meter boxes;

[0043] The meter box identification method selection module determines whether the signal strength attenuation between the meter boxes meets the signal strength attenuation condition and selects the meter box identification method;

[0044] The meter box identification module uses a signal strength attenuation-based identification method to identify the meter box under the branch when the signal strength attenuation between the meter boxes meets the signal strength attenuation condition; and uses a machine learning-based identification method to identify the meter box under the branch when the signal strength attenuation between the meter boxes does not meet the signal strength attenuation condition.

[0045] The meter box sequence identification module identifies the sequence of the meter boxes under each branch based on the results of the branch identification and the meter box identification.

[0046] Furthermore, the branch identification module performs branch identification according to the signal strength attenuation between meter boxes, including:

[0047] When the attenuation value between two meter boxes is ≤ s, the two meter boxes are considered neighbors;

[0048] The set A(A1, A2, ..., A1) of neighbors of each meter box in the statistical area is n ) and the number of elements in each set A set a(a1,a2,……,a n );

[0049] Analyze the set a, accumulate the same number of a in r, r1 represents the number of elements in the set a that are 1, r2 represents the number of elements in the set a that are 2...r 10 Indicates the number of elements in set a that are 10, r 11 Indicates the number of elements in set a that are greater than 10;

[0050] Calculate the ratio of multiple meter boxes:

[0051] rate=(n–r1–r2) / n;

[0052] When rate>r thd When rate≤r thd When s=s+1, repeat the above steps;

[0053] Take the intersection of the elements in each set A and get m sets B (B1B2,…,B m ), set B is the result of branch judgment;

[0054] Among them, n is the number of meter boxes in the area to be identified, s is the attenuation value threshold, r thd It is the empirical value of the ratio of multiple meter boxes.

[0055] According to a third aspect of the present invention, a low-voltage area power grid acquisition system is provided, comprising a master station, a concentrator, and a plurality of smart meters; wherein,

[0056] The master station performs data transmission with the concentrator via a wireless communication network;

[0057] The concentrator includes a CCO, which is connected to the A, B, and C phase power lines and the neutral line of the power line respectively to control the three phases of the AC line and identify the line topology of the low-voltage substation area using the method according to any one of claims 1 to 7;

[0058] Each smart meter includes a STA, and each of the STAs communicates with a CCO in the concentrator to transmit data to the CCO.

[0059] In summary, the present invention provides a low-voltage substation line topology identification method and device, as well as a low-voltage substation power grid acquisition system. In the line topology identification method, the topology is identified based on the attenuation value between two meter boxes. First, branch identification is performed. Then, according to whether the signal strength attenuation between the meter boxes meets the signal strength attenuation condition, a signal strength attenuation-based identification method or a machine learning-based identification method is selected to identify the meter boxes under the branch. Finally, based on the results of the branch identification and meter box identification, the order of the meter boxes under each branch is identified. The identification method and device provided by the present invention can be based on the original low-voltage substation power grid acquisition system. Through calculations using relevant methods, the low-voltage substation line topology identification function can be realized without modifying the hardware system. The identification accuracy is high and the implementation cost is low. Compared with the substation line topology identification scheme in the prior art, it is more suitable for large-scale promotion and application. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a structural diagram of the low-voltage area power grid acquisition system of the present invention;

[0061] Figure 2 It is a flow chart of the low-voltage area line topology identification method of the present invention;

[0062] Figure 3 This is a flow chart of branch identification in the low-voltage area line topology identification method of the present invention;

[0063] Figure 4 It is a schematic diagram of the electrical circuit structure of a certain area;

[0064] Figure 5 It is a flow chart for selecting a meter box identification method in the low-voltage area line topology identification method of the present invention;

[0065] Figure 6 This is a flow chart of identifying the branch meter box based on signal strength attenuation in the low-voltage area line topology identification method of the present invention;

[0066] Figure 7 This is a flow chart of identifying the branch meter box based on machine learning in the low-voltage area line topology identification method of the present invention;

[0067] Figure 8 This is a flow chart of meter box sequence identification in the low-voltage area line topology identification method of the present invention;

[0068] Figure 9 It is a block diagram of the low-voltage substation line topology identification device of the present invention. DETAILED DESCRIPTION

[0069] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention will be further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present invention.

[0070] The technical solution of the present invention is described in detail below with reference to the accompanying drawings. According to one embodiment of the present invention, a low-voltage grid acquisition system is provided, the structural diagram of which is shown in FIG. Figure 1As shown, the system includes a master station, a concentrator, and multiple smart meters. The master station can communicate with the concentrator via a wireless communication network, such as a 4G network. The concentrator includes a central coordinator (CCO) (hereinafter referred to as "CCO"), which is connected to the A, B, and C phases of the power line and the neutral line to control the three phases of the AC line. Each smart meter includes a station (STA) (hereinafter referred to as "STA"), which is connected to any one of the three-phase AC lines and the neutral line. Each STA communicates with the CCO in the concentrator to transmit data to the CCO. The CCO receives two types of data: signal strength attenuation data reported by the STA and other STAs, and voltage, current, and power factor data collected by each smart meter at a specific zero-crossing point (NTB). The system uses the STA to transmit information such as signal strength between meters and voltage and current to the CCO via the power carrier. The CCO then performs relevant algorithm calculations and transmits the results to the master station via 4G. Since the signal strength attenuation between meter boxes is related to the load inside the meter box and the distance between the meter boxes, the present invention proposes to use the above-mentioned collected data to analyze the signal strength attenuation characteristics and perform branch judgment, meter box judgment, and meter box sequence judgment respectively. When processing this information, the CCO processes it separately on a phase line basis, first judging one of the phase lines of phases A, B, and C, and then judging the other phase lines after judging one of the phase lines. Therefore, according to the technical solution provided by the present invention, the topology of the low-voltage substation line can be identified without changing the hardware of the power grid acquisition system.

[0071] According to another embodiment of the present invention, a method for identifying line topology in a low-voltage area is provided. The flow chart of the identification method is as follows: Figure 2 As shown, the following steps are included:

[0072] Based on the signal strength attenuation between meter boxes, branch identification is performed. The flow chart of branch identification is shown in Figure 3. Branch identification can be performed in the following steps:

[0073] First, assume that the number of meter boxes on a phase line in the substation to be judged is n, and s is the attenuation value threshold. Set s to the minimum attenuation value. Start with the minimum attenuation value s, which is the minimum value that can be collected by the attenuation between meters in the entire substation. When the attenuation value between two meter boxes is ≤ s, the two meter boxes are considered neighbors.

[0074] The set A(A1, A2, ..., A1) of neighbors of each meter box in the statistical area is n ) and the number of elements in each set A set a(a1,a2,……,a n); Analyze the set a, accumulate the same number of a in r, r1 represents the number of elements in the set a that are 1, r2 represents the number of elements in the set a that are 2...r 10 Indicates the number of elements in set a that are 10, r 11 Indicates the number of elements in set a that are greater than 10.

[0075] Calculate the ratio of multiple meter boxes:

[0076] rate=(n–r1–r2) / n;

[0077] When rate>r thd When rate≤r thd When s=s+1, repeat the above steps; where r thd is the empirical value of the ratio of multiple meter boxes. Generally speaking, r thd A value of 0.6 can be used. That is, when the proportion of multiple meter boxes is less than or equal to a predetermined empirical value, the attenuation threshold s is increased and identification is repeated. This branch identification process determines whether the standard is met. The standard is met when the number of neighbors of each meter box is relatively large. For example, when the number of neighbors of each meter box exceeds the number of meters in the meter box, the requirement is met.

[0078] Take the intersection of the elements in each set A and get m sets B (B1B2,…,B m ), set B is the result of branch judgment.

[0079] The following is an example of a cell type station area. Figure 4 As shown, Figure 4 This is a schematic diagram of the electrical circuit structure of a certain area. There are two buildings in the area. The wires coming out of the transformer enter the distribution room and then enter the branch box. Four buses are drawn out from the branch box and enter the 3rd to 8th floors and 9th to 11th floors of the two buildings respectively. Figure 4 As can be seen from the diagram, in a building, the meters closest to each other on two branches are the 9th-floor meter and the 3rd-floor meter. However, the distance between the 9th-floor meter and the 3rd-floor meter is much greater than the distance between the meters on the upper and lower floors of the same branch (e.g., the distance between the meters on the 3rd and 4th floors). Assuming the attenuation value of the meters on the 9th and 3rd floors is X, and the attenuation value within the same meter box is x, when the meters with attenuation values ​​greater than x are grouped into a neighbor set, the elements in the set will include the meters on the upper and lower floors of the same branch. Therefore, when the selected attenuation value s is greater than x and less than X, the intersection of each meter's neighbors is the result of the branch. The method for finding the attenuation s is described above and will not be repeated here.

[0080] The flow chart of judging whether the signal strength attenuation between meter boxes meets the signal strength attenuation condition and selecting the meter box identification method according to the judgment result is shown as follows: Figure 5 As shown, you can follow the steps below:

[0081] Assume that the attenuation threshold s = 1. When the attenuation value between two meter boxes is less than or equal to s, the two meter boxes are considered neighbors. The set A (A1, A2, ..., A n ) and the number of elements in each set A set a(a1,a2,……,a n ); Analyze the set a, accumulate the same number of a in r, r1 represents the number of elements in the set a that are 1, r2 represents the number of elements in the set a that are 2... r 10 Indicates the number of elements in set a that are 10, r 11 Indicates the number of elements in set a that are greater than 10.

[0082] If 2*r2 ≤ n–r1, the signal strength attenuation between meter boxes meets the signal strength attenuation condition; if 2*r2 > n–r1, the signal strength attenuation between meter boxes does not meet the signal strength attenuation condition. In other words, when attenuation characteristic analysis reveals that a particular area falls under the category of meter box identification based on signal strength attenuation, machine learning-based meter box identification is employed. This selection of meter box identification methods addresses the issue of inaccurate results from machine learning-based meter box identification when meter box loads are low.

[0083] The method based on signal strength attenuation is used to identify the meter box under the branch. Its basic principle is similar to the branch judgment principle. By finding an attenuation value, under which the number of neighbors of the meter in the meter box is less than the number of meters in the meter box, identification can be made. The flowchart of this method is shown in the figure below. Figure 6 As shown, the following steps can be taken:

[0084] On the premise that the branch result has been identified, s is taken as the minimum attenuation value. When the attenuation value between two meter boxes is ≤ s, the two meter boxes are considered to be neighbors. The set A (A1, A2, ..., A1) of neighbors of each meter box in the station area is counted. n ) and the number of elements in each set A set a(a1,a2,……,a n The judgment of the table box is performed on a branch basis. Under the branch to be judged, the set A of neighbors of each table and the number of elements a in each set are obtained.

[0085] When an element greater than 1 appears in a, proceed to the next step; otherwise, s=s+1 and repeat the above steps.

[0086] Add the decay value threshold s to the decay experience value s j , get the new attenuation threshold s, that is, s=s+s j , s j You can take 4 and repeat the above steps;

[0087] The elements in each set A are intersected to obtain multiple sets B, where set B represents the table box judgment situation under this branch.

[0088] The recognition method based on machine learning is used to identify the branch meter box. It uses the commonly used multivariate linear regression method of the substation line topology recognition to obtain the characteristic value by using the voltage, current and power factor between the two meters at the same time. The flow chart of this method is as follows Figure 7 As shown. The formula for multiple linear regression is as follows:

[0089]

[0090] In the formula, the superscript for voltage U and current I represents time i, and the subscript represents meter 1 or meter 2. X and R represent impedance, and the subscripts R and X represent the real and imaginary currents, respectively. Conventional methods are used to calculate the eigenvalues ​​of each meter pair, and then multiple sets of eigenvalues ​​from the same meter box and different meter boxes are selected as the model. The specific process is that when new voltage, current, and power factor data for two meters are available, multiple linear regression is used to calculate the eigenvalues ​​between the two meters. The Euclidean distance between this set of eigenvalues ​​and each set of eigenvalues ​​in the model is then calculated. These distances are sorted and the three smallest distances are selected. If the three sets contain more eigenvalues ​​from the same meter box, the result is that the meter box is the same, and vice versa. When using machine learning methods to determine the meter box, the eigenvalues ​​of the two meters to be determined are substituted to infer whether the two meters belong to the same meter box.

[0091] Based on the results of the branch identification and meter box identification, the order of the meter boxes under each branch is identified. The flow chart of the identification process is as follows: Figure 8 As shown, the following steps can be taken:

[0092] On the premise that the branch result has been identified, s is taken as the minimum attenuation value. When the attenuation value between two meter boxes is ≤ s, the two meter boxes are considered to be neighbors. The set A (A1, A2, ..., A1) of neighbors of each meter box in the station area is counted. n ) and the number of elements in each set A set a(a1,a2,……,a n ).

[0093] Analyze set A. If a meter box outside the branch appears for the first time in set A, take the meter box corresponding to set A as the first meter box in the branch, mark it as P, and proceed to the next step. Otherwise, set s = s + 1 and repeat the above steps.

[0094] By comparing the attenuation values ​​of the meter box in P and other meter boxes in the branch, find the meter box closest to P and mark it as P;

[0095] If P is the last marked meter box in the branch, then end, otherwise repeat the previous step and continue to find the nearest meter box.

[0096] According to a third embodiment of the present invention, a low-voltage area line topology identification device is provided. The block diagram of the device is as follows: Figure 9 As shown, the device includes a branch identification module, a meter box identification method selection module, a meter box identification module, and a meter box sequence identification module.

[0097] The branch identification module performs branch identification based on the signal strength attenuation between meter boxes.

[0098] The meter box identification method selection module determines whether the signal strength attenuation between meter boxes meets the signal strength attenuation condition and selects the meter box identification method.

[0099] The meter box identification module uses a signal strength attenuation-based identification method to identify the meter box under the branch when the signal strength attenuation between the meter boxes meets the signal strength attenuation condition; when the signal strength attenuation between the meter boxes does not meet the signal strength attenuation condition, a machine learning-based identification method is used to identify the meter box under the branch.

[0100] The meter box sequence identification module identifies the sequence of the meter boxes under each branch based on the results of the branch identification and the meter box identification.

[0101] The specific process of topology identification performed by the above modules is the same as the steps of the identification method provided in the second embodiment of the present invention, and will not be described in detail here.

[0102] In summary, the present invention relates to a low-voltage substation line topology identification method and device, and a low-voltage substation power grid acquisition system. In the line topology identification method, the topology is identified based on the attenuation value between two meter boxes. First, branch identification is performed. Then, according to whether the signal strength attenuation between the meter boxes meets the signal strength attenuation condition, a signal strength attenuation-based identification method or a machine learning-based identification method is selected to identify the meter boxes under the branch. Finally, based on the results of the branch identification and meter box identification, the order of the meter boxes under each branch is identified. The identification method and device provided by the present invention can be based on the original low-voltage substation power grid acquisition system. Through calculation of relevant methods, the low-voltage substation line topology identification function can be realized without modifying the hardware system. The identification accuracy is high and the implementation cost is low. Compared with the substation line topology identification scheme in the prior art, it is more suitable for large-scale promotion and application.

[0103] It should be understood that the above-described specific embodiments of the present invention are merely illustrative or illustrative of the principles of the present invention and do not constitute limitations of the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made without departing from the spirit and scope of the present invention should be included within the scope of protection of the present invention. In addition, the appended claims are intended to cover all variations and modifications that fall within the scope and metes and bounds of the appended claims, or equivalents thereof.

Claims

1. A method for identifying line topology in a low-voltage area, characterized in that: Including steps: Based on the signal strength attenuation between the meter boxes, branch identification is performed, including: when the attenuation value between two meter boxes is ≤ s, the two meter boxes are considered neighbors; s is the attenuation value threshold; The set of neighbors of each meter box in the statistical area is A = (A1, A2, ..., A n ) and the number of elements in each set A is a=(a1,a2,……,a n ); Analyze the set a, accumulate the same number of a in r, r1 represents the number of elements in the set a that are 1, r2 represents the number of elements in the set a that are 2...r 10 Indicates the number of elements in set a that are 10, r 11 Indicates the number of elements in set a that are greater than 10; Determining whether the signal strength attenuation between the meter boxes meets the signal strength attenuation condition includes: setting s=1, calculating r1 and r2; If 2 When r2≤n–r1, the signal strength attenuation between the meter boxes meets the signal strength attenuation condition; if 2 When r2>n–r1, the signal strength attenuation between meter boxes does not meet the signal strength attenuation condition; n is the number of meter boxes in the area to be identified; When the signal strength attenuation between meter boxes meets the signal strength attenuation condition, the identification method based on signal strength attenuation is used to identify the meter box under the branch; when the signal strength attenuation between meter boxes does not meet the signal strength attenuation condition, the identification method based on machine learning is used to identify the meter box under the branch; Based on the results of the branch identification and the meter box identification, the order of the meter boxes under each branch is identified.

2. The method according to claim 1, characterized in that The branch identification according to the signal strength attenuation between the meter boxes further includes: Calculate the ratio of multiple meter boxes: rate = (n – r1 – r2) / n; When rate>r thd When rate≤r thd , s=s+1, and repeat the above steps; Take the intersection of the elements in each set A and get m sets B=(B1B2,…,B m ), set B is the result of branch identification; Among them, r thd It is the empirical value of the ratio of multiple meter boxes.

3. The method according to claim 2, characterized in that The identification method based on signal strength attenuation is used to identify the branch meter box, including: Obtain the set A of neighbors of each table box under the branch to be identified and the set a of the number of elements in each set A; When an element greater than 1 appears in a, proceed to the next step; otherwise, s=s+1, and repeat the steps in claims 1-2 above; s=s+s j , and repeat the steps in claims 1-2 above; Intersect the elements in each set A to obtain multiple sets B, where set B represents the identification of the table box under this branch; Among them, s j To decay experience value.

4. The method according to claim 3, characterized in that The method of identifying the branch meter box using a machine learning-based identification method includes: Obtain the eigenvalues ​​of the multivariate linear regression formula between two electricity meters; Select multiple sets of characteristic values ​​of the same meter and different meters as models; Calculate the Euclidean distance between the eigenvalue and each set of eigenvalues ​​in the model; Sort the Euclidean distances and take the three smallest ones. When there are many groups of characteristic values ​​of the same box represented by the three Euclidean distances, the result to be identified is the same box.

5. The method according to claim 4, characterized in that The multiple linear regression formula includes the following formula: 。 6. The method according to claim 5, characterized in that The identification of the order of the meter boxes under each branch includes: Get the set A of each table box under this branch; If a meter box outside the branch appears for the first time in set A, take the meter box corresponding to set A as the first meter box in the branch, mark it as P, and proceed to the next step; otherwise, s=s+1, and repeat the above steps; By comparing the attenuation values ​​of the meter box in P and other meter boxes in the branch, find the meter box closest to P and mark it as P; If P is the last marked table box in the branch, then end, otherwise repeat the previous step.

7. A low-voltage area line topology identification device, characterized in that: It includes a branch identification module, a meter box identification method selection module, a meter box identification module, and a meter box sequence identification module; wherein, The branch identification module performs branch identification based on the signal strength attenuation between the meter boxes, including: when the attenuation value between two meter boxes is ≤ s, the two meter boxes are considered to be neighbors; s is the attenuation value threshold; the set of neighbors of each meter box in the station area is counted as A = (A1, A2, ..., A n ) and the number of elements in each set A is a=(a1,a2,……,a n ); Analyze the set a, accumulate the same number of a in r, r1 represents the number of elements in the set a that are 1, r2 represents the number of elements in the set a that are 2...r 10 Indicates the number of elements in set a that are 10, r 11 Indicates the number of elements in set a that are greater than 10; The meter box identification method selection module determines whether the signal strength attenuation between the meter boxes meets the signal strength attenuation condition and selects the meter box identification method, including: setting s=1, calculating r1 and r2; If 2 When r2≤n–r1, the signal strength attenuation between the meter boxes meets the signal strength attenuation condition; if 2 When r2>n–r1, the signal strength attenuation between meter boxes does not meet the signal strength attenuation condition; n is the number of meter boxes in the area to be identified; The meter box identification module uses a signal strength attenuation-based identification method to identify the meter box under the branch when the signal strength attenuation between the meter boxes meets the signal strength attenuation condition; and uses a machine learning-based identification method to identify the meter box under the branch when the signal strength attenuation between the meter boxes does not meet the signal strength attenuation condition. The meter box sequence identification module identifies the sequence of the meter boxes under each branch based on the results of the branch identification and the meter box identification.

8. The device according to claim 7, characterized in that The branch identification module performs branch identification according to the signal strength attenuation between the meter boxes, and further includes: Calculate the ratio of multiple meter boxes: rate = (n – r1 – r2) / n; When rate>r thd When rate≤r thd , s=s+1, and repeat the above steps; Take the intersection of the elements in each set A and get m sets B=(B1B2,…,B m ), set B is the result of branch identification; Among them, s is the attenuation threshold, r thd It is the empirical value of the ratio of multiple meter boxes.

9. A low-voltage power grid acquisition system, characterized in that: It includes a master station, a concentrator, and multiple smart meters; among them, The master station performs data transmission with the concentrator via a wireless communication network; The concentrator includes a CCO, which is connected to the A, B, and C phase power lines and the neutral line of the power line respectively to control the three phases of the AC line and identify the line topology of the low-voltage substation area using the method according to any one of claims 1 to 6; Each smart meter includes a STA, and each of the STAs communicates with a CCO in the concentrator to transmit data to the CCO.

Images